Benchmark: Randomised benchmarking family
Benchmarks / Component level (gates and operations)
The average error rate of a set of gates, under a general noise model including time- and gate-dependent errors.
- Level
- Component level (gates and operations)
- Maintained by
- A family of protocols from the research literature, implemented in most vendor toolkits.
- European
- No
- Reference
- https://arxiv.org/abs/1009.3639
How it is measured
Random sequences of gates of increasing length that ideally compose to the identity are run; the decay of the success probability with length gives the average error per gate (Magesan et al. 2011). Variants include interleaved, simultaneous and direct randomised benchmarking.
Limitations and criticisms
- An average over gates and noise; it can hide coherent and correlated errors that matter for algorithms.
Used in European procurements
- Used in: EuroHPC procurement EuroSSQ-HPC (Required in the benchmark suite (scored under the award criteria))
As the tender documents state; how a benchmark is weighted against others is described in each procurement.
Sources
- E. Magesan, J. M. Gambetta, J. Emerson (2011). Robust randomized benchmarking of quantum processes. Phys. Rev. Lett. 106, 180504 (2011); arXiv:1009.3639. Read: abstract page, retrieved 2026-10-23.